作者: Catherine Evans
DOI:
关键词:
摘要: This thesis addresses the need for a suitable methodology to populate geometallurgical models of an ore deposit with recovery values valuable mineral(s) present in ore. The approach taken this research was identify whether processing characteristics ores, case target mineral flotation, could be estimated from quantitative measurements microtexture made on coarse particulate samples crushed If technique feasible it would have benefit reducing amount physical flotation testing required modelling programmes, together corresponding reductions analysis costs, turnaround time and sample mass requirements. overall developed is key contribution work comprises three sequential stages which generate estimate elemental each that can used as input deposit. are: − characterisation using samples, simulating breakage means ranking liberation separability response ores based surface composition particles created when broken feed size distribution kinetic minerals measured under standard conditions. In developing hypotheses central were tested, these being that: affect their by behaviour given class does not vary spatial location batch rougher model calculates particle surfaces data. research, four one drill hole Cadia East Cu-Au porphyry located near Orange, New South Wales, development further wider range locations demonstrate application methodology. A method estimates coefficient variation microtextural such grain size; method, bootstrap resampling technique, also most appropriate use well number must achieve level error user. provides confidence limits data distributions. notable finding that, analysed work, chalcopyrite had greater effect than copper head grade after grinding. All studied shown break non-randomly ground distribution. Thus simulated imposing random geometric pattern onto images texture, absolute value accurately predicted simulations; all cases observed higher value. However, relative correct some support first hypothesis. applying group similar microtextures; representative archetypes then physically provide accurate information predicting members group. An investigation showed while responses limited significantly different, similarity did extend body. These more-widely contained significant proportions exhibited slower kinetics single hole; deeper regions proportion identified non-floating since recovered test Therefore second hypothesis supported published predict data, results predicts where only fast-floating form; third supported. For there clear relationship classes between rate surface, increasing increased. contains slow-floating and/or forms over-estimates therefore ores. Further account presence fast-floating, mineral. proposed here practical rank separation form has rate.